1. Field of the Invention
The present invention generally relates to the transmission of digital video data signals and more particularly, is applicable to data transmission in a digital video tape recorder (D-VTR) or the like which digitally records and reproduces high definition television (HDTV) video signals.
2. Description of the Prior Art
High definition television (HDTV) techniques are known for enhancing the resolution of the video image by widening the transmission band relative to that of the standard NTSC (National Television System Committee) or similar system and thereby achieving improved image quality. Telecasting based on such HDTV techniques is currently under study along with development of digital video tape recorders (D-VTR) capable of recording and reproducing HDTV video signals in a digitized form.
In a D-VTR which digitally records and reproduces HDTV video signals having a transmission band far wider than that of the standard NTSC video signals, a great amount of video data has to be processed at a high speed, and it is difficult, with existing devices and circuit technology, to provide a circuit arrangement that is adequate for directly processing such video signals. Therefore, it has been proposed heretofore to adopt a data processing method in which the processing speed is reduced by distributing the video data into number of channels for performing parallel operations thereon. For example, in a publication entitled "An Experimental HDTV Digital VTR With A Bit Rate of 1.188 Gbps", authored by persons having an obligation to assign to the assignee of the present application and appearing in "IEEE Transactions on Broadcasting" December, 1987, Vol. BC-33, No. 4, at pages 203-209, and also in U.S. patent application Ser. No. 06/925,167, filed Oct. 31, 1986, and which has been replaced by Continuation application Ser. No. 07/213,442, filed June 28, 1988, and each having a common assignee herewith, a scheme for effecting parallel data transmission is shown to involve division of the HDTV picture, at vertically extending lines of division, into N horizontally contiguous segments of the picture, whereupon the horizontal time axis of each segment is expanded, as in a memory, so as to reduce the signal processing speed for each picture segment. By using such a picture division method, the correlation between picture elements (pixels) can be fully utilized, that is, a digital filter can be provided for each divided picture so as to permit error concealment in the event of errors lying outside the error correcting ability of circuits provided therefor. More specifically, in the signal processing system being here described, the luminance signal Y and two chrominance signals R-Y and B-Y are individually sampled to provide a 4:2:2 signal format, that is, every other R-Y and B-Y signal output is dropped or the chrominance signals are sub-sampled at a lower sampling frequency than the luminance signal, whereupon the Y and remaining R-Y and B-Y signals for each segment are multiplexed, coded for error corrections and then converted to serial form for recording. In such case, most of the signal processing takes place in an 8-channel format.
In a video tape recorder (VTR), it is generally impossible to record or reproduce DC or low frequency components by means of the magnetic head used for recording and reproducing and, when such head is rotary and electrically coupled to the corresponding recording and reproducing circuits via a rotary transformer, difficulties are also encountered in transmitting the low frequency components to and from the rotary head. Furthermore, since the high frequency characteristic is deteriorated by the spacing loss, head gap loss and so forth in a magnetic recording operation, the recording/reproducing circuit of the VTR has a band-pass type frequency characteristic.
Therefore, in existing D-VTRs designed to perform digital signal recording and reproducing operations, various recording-modulation encoding methods are relied upon for conversion of the digital signals into a form substantially suited for the characteristic of the magnetic recording circuit. In accordance with one such recording-modulation encoding method, m-bit data representing one sampled or picture element (pixel) is converted into n-bit recording data, and a series of such recording data are converted by a non-return to zero (NRZ) or similar mode for recording in that mode. One of such recording-modulation encoding methods employs an 8-10 conversion mode which converts 8-bit data into 10-bit data having a satisfactory code balance for minimizing the DC component of the recorded signals. However, such 8-10 conversion undesirably increases the bits of data to be recorded, that is redundant bits have to be recorded. In order to avoid that disadvantage, an 8-8 conversion mode has been employed which rearranges the 8-bit video data without increasing the bit number thereof by utilizing the close correlationship between adjacent samples of the video signals. Although low-frequency components of the recording signals are reduced without increasing the bits of recorded data, the 8-8 conversion mode is limited in its ability to eliminate the low-frequency components of the recording signals.
Furthermore, in the existing recording/reproducing circuitry or unit in a D-VTR, an error correction code of a product code pattern is added to the digital data to be transmitted, with the intention that any data error, such as is caused by data dropout or the like, is detected and corrected or concealed at the receiving or reproducing side.
In the earlier described known D-VTR in which each image frame composed of HDTV video signals is divided into horizontally contiguous N segments to reduce the video data processing speed to 1/N, the video data of the divided frame segments are distributed into respective channels for parallel processing. In other words, the video data of each of the frame segments is distributed to a corresponding individual channel. Consequently, if any fault or increase of the error rate occurs in one channel, it becomes necessary to frequently execute error correction of the reproduced data in a manner consistent with the increase of the error rate in such channel. Moreover, if the video data of a channel is not reproducible at all, error correction becomes impossible and eventually the divided frame segment corresponding to that channel cannot be reproduced.
Further, it has been proposed, for example, as disclosed in an article entitled "Study of Multi-channel Distribution for High Definition Digital VTR" by Shinichi Mayazaki, Yoshizumi Etoh and Masuo Umemoto, appearing at pages 229-230, of the 1985- Proceedings of the Institute of Television Engineers of Japan, to adopt run length limitation 8-8 conversion and to use two pixels as the unit of distribution of the luminance and chrominance signals which are sequentially circulated in each of a number of channels. The channel receiving the signals is shifted from one to another at every horizontal period so that the data distributed into any one of the channels does not correspond to vertically aligned pixels of the reproduced picture. The foregoing has the advantage of permitting the use of vertically adjacent data for concealment of continuous errors in a channel due to drop out or the like. However, the described multi-channel distribution for a high definition digital VTR, does not substantially eliminate low frequency components of the recording signals by its use of the run length limitation 8-8 conversion. Further, since the high speed digital video data signals are not spatially divided into a number of horizontally contiguous segments of the video picture, the described multi-channel distribution does not adequately reduce the data bit rate for processing in each channel.